The Midterms: 'Blue wave' vs 'thin red line'

Both Democrats and Republicans face an uphill battle to achieve their objectives in the 2018 midterms.

On the one hand…

This is traditionally the time in America’s political cycle when the pundits start to look forward to the midterm elections, next due to be held on 6 November 2018. This year is proving no exception; indeed, if anything, interest and speculation are greater than ever, spurred in particular by the Republican Party’s loss of a supposedly rock solid 'red' Senate seat in Alabama earlie r this month.

If this is hardly surprising, it is no less so that, putting the ‘on the one hand, on the other hand’ school of thought to one side, opinion currently seems to be divided between those who think the Democrats are firmly on track to win a majority in one or both Houses and those who are forecasting that the Republicans will (or at least should) hang on to their majorities.

What I shall try to do in this article is summarise the case underpinning the two schools of thought, not in the hope of reaching a reliable forecast but as a guide as to what pointers we should follow in particular in the coming weeks to try to glean a clearer idea of the likely ultimate outcome. In so doing, let me put both my hands up at the outset and confess that I am personally (and somewhat uncharacteristically) in the ‘hedging my bets’ camp, at least for now!

What’s at stake?

In posing the question in the heading of this section, let me be clear that I do not mean what’s at stake in terms of either likely policy shifts or even President Donald Trump’s prospects for surviving his full four-year term. These are undoubtedly very important topics; but they are not the focus here. Rather, I shall run briefly through the main features of what the elections per se are all about. I appreciate that many readers will be familiar with what follows in this section (and those who are should feel free to skip over it); but others may not be and it is important contextually, in my view.

In the Senate, all 33 Class 1 seats ((out of 100 seats in total) are up for grabs. The winners will serve a six-year term starting 3 January 2019. A full list of the states in play can be found here. Key points to keep in mind are as follows:

  • The Democrats will be defending 23 seats Class 1 plus two held by Independents who caucus with them;
  • Furthermore, they will have to defend one Class 2 seat in a special election in Minnesota following the resignation of incumbent Al Franken and the appointment of Tina Smith (who is expected to stand);
  • Of these, Mr Trump won ten in last year’s general election, underlining that they are, in theory at least, vulnerable (more on which below);
  • The Republicans have eight seats to defend, only one of which is in a state won by Hillary Clinton last year;
  • However, the Republicans will lose the (well-documented) advantages associated with incumbency in Arizona and Tennessee thanks to the already announced retirements of, respectively, Jeff Flake and Bob Corker.

The Republicans currently hold a 51-49 advantage in the Senate (ie following their loss in Alabama). So (with Vice President Mike Pence holding the casting vote), the Democrats require a net gain of two seats to win control of the Senate.

In the House of Representatives, all 435 congressional districts are up for grabs across the 50 US states. The winners will serve in the 116th US Congress from 3 January 2019 until 3 January 2021.

  • The Republicans won 241 seats in last year’s election (a drop of six relative to the 2014 Mid-terms);
  • Of those, 23 are in districts won by Mrs Clinton in 2016;
  • The Democrats won 194 seats in 2016;
  • Of those, ten are in districts won by Mr Trump;
  • Expert opinion has it that only around 50 seats are genuinely contestable in ‘average’ circumstances.

One way or the other, the Democrats require a net gain of 24 seats to win a majority in the House.

[Note: 6 November 2018 will also see gubernatorial elections in 36 US states and three US territories. Furthermore, 87 out of 99 state legislature chambers are also due to hold elections. These ballots are not only important in their own right but also because many states task governors and state legislators with drawing up congressional district boundaries, with a review of the current boundaries due in the wake of the 2020 national census.]

The ‘blue wave’

Although we have learned in the past couple of years to be very wary of opinion polls, in weighing the prospects of each party it is still worth starting with a consideration or national polling for the proverbial ‘generic’ Democrat vs Republican which, at this stage in the electoral cycle, has historically been a pretty good indicator.

The RCP average (which has long been been and remains one of my favoured indicators) gives the Democrats a whacking13 percentage point lead based on polls taken between 4 and 19 December. Interestingly, it also puts the Republican Party at 36.1%, ie within the 36-38% band which many polls taken over the past six months or so have indicated is Mr Trump’s personal approval rating (ie an all-time low for this stage in a presidential term of office).

[Note: the RCP average currently puts Mr Trump’s approval rating at 38.6%.]

A statistical analysis by respected political scientist Alan Abramowitz concludes that:

“If Democrats maintain a [generic] lead in the high single digits, that probably indicates they will have a decent chance to win the House or at least significantly cut into the Republicans’ majority. A bigger Democratic lead, into the double digits, would make a takeover more likely.”

Second, there are compelling indicators favouring the Democrats from special elections which have been held since the start of the year. A 13 December (ie post-Alabama) analysis by Harry Enten of FiveThirtyEight concludes as follows:

“…if you’re a Democrat, Tuesday’s Alabama result is just the latest special election sign that things are looking up heading into 2018”.

His analysis of over 70 elections held so far for federal and state legislature seats shows that the Democrats have been performing 12 percentage points better on average than the ‘partisan lean’ (ie how a state or district would be expected to vote in a nationally 50:50 presidential election) would lead one to expect; and that they have out-performed in 74% of these elections (and 100% of the seven federal elections!). This doesn’t mean that the Democrats have won 74% — far from it, in fact; rather:

“The point is that Democrats are doing better in all types of districts with all types of candidates. You don’t see this type of consistent outperformance unless there’s an overriding pro-Democratic national factor”.

Mr Enten goes on to show how the average swing in special federal elections has been a pretty good indicator of the midterm outcome since 1994. However, he also underlines that, with just six data points here, “we shouldn’t get carried away drawing conclusions”.

Third, one very important reason for the Democrats’ outperformance has been the energising of their vote, opening up what, among others, Nate Silver of FiveThirtyEight refers to as an enthusiasm gap between Democrat and Republican voters. This was particular clear in Alabama this month and in November’s elections in the state of Virginia. An article in the 19 December edition of The Economist (subscriber access only) offered the following consideration:

“Democrats appear more motivated than Republicans. Turnout surged in liberal northern Virginia, and the big cities and black-majority counties of Alabama. In Alabama turnout fell in majority-white, rural areas; that was not as true in Virginia, but Democrats banked enough votes in urban centres to cancel out the Republican showing in rural areas.

In both states, non-white and young voters broke decisively for Democrats. [Democrat] Mr Jones won 96% of black voters—and 98% of black women, while [Democrat gubernatorial candidate] Mr Northam took 80% of Virginia’s non-white voters. Both candidates won majorities of voters younger than 45, decisively lost voters 65 and older and barely lost voters between the ages of 45 and 65. That should worry Republicans: Americans born since 1980 have taken over from baby-boomers as America’s largest generation, and 43% of millennial adults are non-white. Republicans are appealing to a dying generation at the expense of a growing one.

Republicans should also worry about their slipping hold on the suburbs. Mr Jones decisively won Alabama’s five biggest cities and their surrounding counties, three of which Mr Trump won by 13 points or more.”

Fourth, ‘Bannonisation’. Mr Trump’s former Chief Strategist, Steve Bannon, has shown no sign as yet that, following the defeat in Alabama of Roy Moore whom he backed heavily, he is about to step away from his commitment to put up like-minded candidates in Republican primaries where he deems the incumbent to be insufficiently ‘Trumpian’ in her or his world view. This may be justified in one respect, ie that Mr Moore was personally just too toxic rather than too conservative; but any candidate that conservative is going to struggle to win over middle ground voters in marginal seats where there is a credible Democrat standing. Indeed, given the traditional incumbency advantage, damage to the Republicans’ chances may have been done already with an unusually high 25 House incumbents indicating that they are not going to run for re-election, some certainly because they see themselves as vulnerable to a Bannon-backed insurgent.

[Note: 14 House Democrats have also indicated that they will be retiring; but this is fairly average for this stage in the cycle.]

The ‘thin red line’

First, generic polling. The bottom line in the article in The Economist I referenced earlier was that, despite all the favourable data, not least generic polling, the Democrats still have a mountain to climb to win a majority in one house, let alone both. As the article puts it:

“Data such as these suggest that the midterms of 2018 could be a wave election for Democrats — but thanks to gerrymandering and their voters’ concentration in urban centres, translating enthusiasm into congressional majorities may prove difficult.”

The Economist also underlines that in the Senate the requisite net gain of two seats means that the Democrats have to defend successfully all 24 of the seats they currently hold which are up for grabs and win two seats held by the Republicans. So, caution over their prospects there seems to me to be entirely called for.

Second, special elections. Mr Enten’s comparison in his analysis which I referenced earlier between today and 2006, when the Democrats gained 30 seats to take control of the House, is all well and good. But the consensus is that electoral district boundary changes brought in sin e then tend on balance to favour the Republicans, thereby making the Democrats’ task even harder in 2018 than the one they faced 12 years previously.

Third, the enthusiasm gap. Let’s start by keeping firmly in mind that pretty much all the data accumulated so far since Mr Trump was sworn in predates the passing of tax reform legislation last week. As things stand, only about one-third of Americans have been telling pollsters that they support the bill (ie again more or less consistent with the President’s personal approval rating!), one might imagine that its passage would do little to boost voters’ enthusiasm for Mr Trump and the Republican Party. However, ‘Obamacare’ faced similar opposition when it first passed into law and yet now enjoys majority support. Nevertheless, the shift in public sentiment did not come in time to save the Democrats in the 2010 midterms. The Republicans clearly face a similar challenge (and seem to be adopting similar tactics to ‘sell’ the tax bill); but they may have an advantage relative to the Democrats in 2010 if it is indeed the case that most Americans will see at least a modest boost (the Tax Policy Center calculates a 2.2% increase on average) to their take-home pay next year.

Furthermore, Republicans’ fortunes in the midterms are likely to be linked to some extent at least to Mr Trump’s personal approval ratings. History suggests that these could improve over the coming months (or they could, of course, slide further). We may be sure that the President will do everything he can to shore up and boost his ratings. Indeed, there are increasingly clear signs that he is setting the stage to deliver on at least some of his pre-election commitments on trade, which would play strongly with his base including white working class voters whom the Democrats really need to win back from him.

Fourth, ‘Sandersisation’. Although it is a fundamental rule of politics that governments tend to lose elections rather than oppositions winning them, the Democrats still need to field credible candidates. As it is, the party’s left-liberal wing, personified by Bernie Sanders and Elizabeth Warren, remains determined to pursue a ‘progressive’ agenda which could yet undermine the Democrats’ midterm prospects; and it seems currently to hold the whip-hand over the centre-moderates.

What to look out for

Despite the need for some caution it would be unwise to ignore entirely the generic national opinion polls. In this respect especially, I find the RealClearPolitics website a good source. However, let’s put those to one side for now and consider other potentially useful indicators which may prove to be more reliable.

First, special elections. There will be at least three for the House between now and 6 November, ie:

  • Pennsylvania’s 18th Congressional district to be held on 13 March and won in 2016 by the Republicans 54-46%;
  • Arizona’s 8th Congressional District to be held on 24 April and won in 2016 by the Republicans with 69% in a two-horse race with a Green candidate; and,
  • Ohio’s 12th Congressional District, date to be announced, and won in 2016 by the Republicans with 67% of the vote.

In terms of actual ‘wins’, probably the best the Democrats could expect would be one from three. But these are still well worth watching to see if they can continue to out-perform significantly the ‘partisan lean’.

Second, the primaries. With the exception of Texas, which holds its primaries in March, the season runs from April to September. The key point here with both parties is to look out of insurgent candidates ousting moderate incumbents.

Third, states to watch closely in the Senate race. The Democrats are expected to target in particular Arizona and Nevada; and they may also pitch hard at Tennessee, Texas and Utah. The Republicans are spoiled for choice with, by some reckonings, 13 seats they could flip; but their best bets look to be Indiana, Missouri, Montana, North Dakota and West Virginia, all of which were won not only by Mr Trump last year but also by Mitt Romney in the 2012 general election.

Fourth, districts to watch closely in the House race. A useful list of the Republican-held districts which the Democrats most need to flip can be found here. However, they are also likely to have to fight defensive battles in at least the 12 districts they hold at present but which were won by Mr Trump in 2016.

Fifth, Sabato’s Crystal Ball (or similar). This is one of a number of services (one of its pluses being that it is free!) which offers a decent breakdown of all contested seats into one of six categories, ie looking at the seats of the two parties in turn and dividing them into ‘likely’, ‘leans’ and ‘toss-up’.


History and current data seem to suggest that the Republicans may face a bigger challenge holding on to their majorities than the Democrats do in winning back control of at least one house if not two. However, neither history nor the polls may be a reliable guide in 2018, any more than they were in the 2016 presidential election. Furthermore, the proverbial ‘events’ could yet blow either party off course. So, for now I am going to continue to sit on the fence while following my chosen indicators very closely indeed in the hope of a clearer idea of the likely outcome emerging in the not too distant future.

Alastair Newton